2020 Presidential Election

[Meadmaker]

Here's what has me worried. All the probability calculations assume no systemic error in polling. If there is a systemic error, it's in favor of Biden.

So, playing with 270towin's interactive map, I start by saying that all the toss up states go to Trump. (The big prizes, with the most uncertainty, are Ohio and Florida.) So, Trump wins those, and that leaves Biden with 290.

That means he can still afford to lose Pennsylvania, plus any other 1 electoral vote.

I don't think it's in the bag for Biden yet.

 

[acbytesla]

Here's what has me worried. All the probability calculations assume no systemic error in polling. If there is a systemic error, it's in favor of Biden.
So, playing with 270towin's interactive map, I start by saying that all the toss up states go to Trump. (The big prizes, with the most uncertainty, are Ohio and Florida.) So, Trump wins those, and that leaves Biden with 290.

That means he can still afford to lose Pennsylvania, plus any other 1 electoral vote.

I don't think it's in the bag for Biden yet.

How do you come to that conclusion? Maybe a systemic error favors Trump. I mean if you don't know if there is one how can you say who it favors? Sort of a fallacy don't you think?

 

[dirtywick]

Nobody really talks about it as an option for some reason but they may be overcorrecting in their polling method changes to avoid another 2016 mistake.

There's a lot to look at with polls, and there's no question Trump is fairly popular to the tune of tens of millions of supporters. But if there's secret shy voters, it doesn't explain the massive fundraising gap. It's hard to explain that away imo

 

[acbytesla]

Nobody really talks about it as an option for some reason but they may be overcorrecting in their polling method changes to avoid another 2016 mistake.

There's a lot to look at with polls, and there's no question Trump is fairly popular to the tune of tens of millions of supporters. But if there's secret shy voters, it doesn't explain the massive fundraising gap. It's hard to explain that away imo

Trump supporters hardly seem shy. They wear hats, wave huge Trump flags and threaten people with guns. Shy? I don't think so.

 

[Meadmaker]

How do you come to that conclusion? Maybe a systemic error favors Trump. I mean if you don't know if there is one how can you say who it favors? Sort of a fallacy don't you think?

I think the biggest source of systemic error is failure to account for people who refuse to talk to pollsters. I think polls are just another element of "the media" that Trump has vilified for four years/

I read a story about polling yesterday that said a huge problem in 2016 polls was failure to account for education in their polling samples. They would make polls based on matching poll respondents based on age, race, and income, but not based on education. So, they oversampled people with college degrees, because people with college degrees are more likely to talk to pollsters. This year, they are making sure to include non-college educated people, and they think it's correcting the problem.

I don't do polls for a living, so I can't be confident that I'm right. That's just what I think. I think people are even less likely to talk to pollsters than they were four years ago, and I think that Trump supporters are less likely to talk to pollsters than Biden supporters, and I don't think the pollsters are adequately compensating.

So, if I had to make a prediction, I would take all the toss up states and throw them to Trump. On 270towin's map, that leaves Biden with 290, including 20 in Pennsylvania. That's good enough for a win, but it doesn't inspire a lot of confidence. There's not much margin for error.

 

[The Atheist]

So, if I had to make a prediction, I would take all the toss up states and throw them to Trump.

Ah, you're an old pessimist. Better to live in hope than confidence?

Give confidence a try - it'll feel even better when you watch the board turn blue.

 

[Dr.Sid]

Ah, you're an old pessimist. Better to live in hope than confidence?

Give confidence a try - it'll feel even better when you watch the board turn blue.

Optimism, sure. Hope, fine. But don't call it win till it's over. 2016 should be lesson to everyone. Poll are polls. Elections are elections.

 

[angrysoba]

Apparently there's a poll that shows the orange ghoul ahead in Pennsylvania and Michigan?

When I looked on 538, the only one I could see was by a polling company called The Trafalgar Group. They get a C- from 538, so it is probably nothing to worry about right now.

 

[llwyd]

Or.......there would have to be some systemic error in the way the polls were conducted that consistently favored Hillary. The pollsters didn't think that was worth worrying about, but it turns out they were wrong. That was the great revelation of 2016.

There was a great focus on national polls which constantly showed Hillary ahead I think by 2-4% by the polling day. And as it turned out she really was ahead and those figures were actually pretty reliable. What the polls missed were some battleground states like Pennsylvania and Michigan which had the reputation of being like Texas, looking tight but always going Democratic in the end. So, the pollsters were not as much wrong as in the popular imagination.

 

[SezMe]

Trump Tweets

For years you had a President who apologized for America—now you have a President who is standing up for America! Get your friends, get your family, get your neighbors and GET OUT & VOTE! On 11/3 we must finish the job and Drain the Swamp once and for all!

Because I'm gonna be prez for the rest of the life of the country. Like those Kims in NK, I'll rule even after I'm on the wrong side of the grass.

 

[SezMe]

EXCLUSIVE: That's not exclusive. <snip>

"Exclusive" is the most over used and abused word in the media. The worst example I know of is when some high-level administration official makes the rounds of the Sunday morning news shows, at least one channel will claim, "EXCLUSIVE!!!! interview with <someone>. Sheeeesh.

 

[Ambrosia]

Looks like President Trump is going to hold Florida easily.

Are alarm bells ringing in Democratic Party HQs across the land.

Has this all been a blue mirage ?

Will President Trump get the EC and the popular vote this time around ?

Highly unlikely.

538 shows odds of about 3% that Trump wins both EC and the popular vote.

Their forecast shows him with a 10% chance of re-election as of the time of this post.

He's not winning this one. He might scrape a win in Florida, but that's not enough.

 

[SezMe]

Actually, if he goes back to a private citizen and continues to tweet as he has been, I'd imagine he'd likely be banned sooner or later. Being president has certain perks, for reasons I don't fully agree with.

Because he claims his tweets are official government correspondence. That gives him carte blanche. The most that can be done is to put a label on them and, as we have seen, that drives Trump and the right wing into paroxysms over "censorship".

 

[SezMe]

We're just in bizarre territory. I hope what it means is that there will be massive voter turnout overall.

And I'm keeping an eye on weather maps. It's possible that Hurricane Eta will be making landfall on election day. (There aren't any official forecast tracks yet, and it isn't certain the the current system will develop into a named storm, but it's looking more likely. It might not make landfall by Tuesday though, and maybe not in the US.)

Because Trump dropped his Sharpie under his desk and he's too fat to reach down to get it.

 

[SezMe]

He'll demand that they stop counting instead.

AIUI the voting machines can tally automatically AND because there are proportionally fewer votes, the counts from these will be available more or less immediately. OTOH, postal votes will need to be counted manually, in many states counting cannot start until election day (and is expected to continue past it), and the orders of magnitude larger number of votes mean that these processes will be overwhelmed.

Look for the tally for a state to be the in person votes, 25% of the mail in votes and as few of the early votes as they can get away with.

I don't think that is true. The mail in votes have to be verified and that's what takes time. After that step they're just fed into a scanner like all the other ballots.

 

[The Don]

Highly unlikely.

538 shows odds of about 3% that Trump wins both EC and the popular vote.

Their forecast shows him with a 10% chance of re-election as of the time of this post.

He's not winning this one. He might scrape a win in Florida, but that's not enough.

I think that whilst fivethirtyeight are doing their best to reflect the polls, and if the polls were accurate then their forecast will be accurate, their forecast is vastly underestimating President Trump's probability of victory.

I think that in swing states there are a *LOT* of Republicans, and Republican-leaning independents who are either tight-lipped or who are telling pollsters that they are not voting for President Trump who will end up voting for him on election day. They're split between those people who genuinely think they will vote for Biden but will switch at the last minute, and those who don't want to broadcast their support for President Trump.

I also think that the Republicans will be very effective at challenging and overturning tight results in swing states and I don't think fivethirtyeight's forecast can reasonably take into account post-election activity like this. President Trump can lose both the popular vote and Electoral College and still win the election.

 

[The Don]

I don't think that is true. The mail in votes have to be verified and that's what takes time. After that step they're just fed into a scanner like all the other ballots.

OK thanks

If mail in votes take longer to count - because of the verification process, the GOP have an opportunity to stop the count before all mail in votes have been tallied. I also think that they will try to make sure that as few of the verifications are successful as possible.

 

[SezMe]

He's the Executive of the Federal Government. He has no authority over the state governments, their policies, their infrastructure.

Well, that's not true, is it. The Executive Branch, via the EPA, sets water quality standards across the USA. Another group (I forget which) sets policy on all navigable waterways and the river from which Flint drew their water was a navigable waterway.

The actual situation is not nearly as simple as you are attempting to make it.

 

[SezMe]

For our non-US members, the first results will start coming in around the early morning Zulu hours about 0130. Nobody freak out when the early results favor Trump. The smaller, rural precincts always go to the Republican and they report first. The larger suburban and urban precincts tend to favor Democrats and report their results last.

There's some district in Maine that has only a few people - maybe a dozen? So they all vote right after midnight, count their votes, and report in just so they can be the first . Doesn't mean squat in Maine, much less the nation.

 

[SezMe]

That percentage seems implausibly high. The nation is strongly and evenly divided, in the last election. This seems like predicting a number of close elections will break Biden's way.

I can see guessing that Florida is likely to go to Biden. But guessing that there's a 90% chance that all of the swing states will go to Biden seems a bit much.

Biden is polling ahead in all swing states but Ohio. It's not an unreasonable prediction. This election is *NOT* a rerun of 2016; that's where your error is.